Amphibia: Gymnophiona), and a Comparison to Dermal Ossifications of Other Vertebrates
Total Page:16
File Type:pdf, Size:1020Kb
.I I ’.i /’ . , JOURNAL OF MORPHOLOGY 2062-43 (1990) Structure of the Scales of Dermophis and Microcaecika (Amphibia: Gymnophiona), and a Comparison to Dermal Ossifications of Other Vertebrates LOUISE ZYLBERBERG AND MARVALEE H. WAKE CNRS UA 161 and Laboratoire d’Anatomie Comparhe, Universith de Paris VII, Paris 75251, France (L.Z.);Department of Integrative Biology and Museum of VertebrateZoology, University of California,Berkeley, California 94720 ABSTRACT The structures of the dermal scales and the cells surrounding the scales in two species of gymnophione amphibians were studied using histochemistry and light, scanning and transmission electron microscopy. Scales are composed of a basal plate of several layers of unmineralized collagenous fibers topped with mineral- ized squamulae. Squamulae are composed of numerous mineralized globules and mineralized, thick collagen fibers. Mineralization is therefore both spheritic and inotropic. Isolated flattened cells lie on the outer surface of the squamulae and seem to be involved in mineral deposition. Cells that line the basal plate synthesize the collagenous stroma of the plate. Each scale lies in a thin connective tissue pocket, and a large connective tissue pouch includes several scales in each annulus. The similarities of gymnophione scales to elasmoid scales of osteichthyans are largely superficial. Aspects of mineralization and of pocket development differ considerably. There are also similarities, as well as differences, in the gymnophione scales and osteoderms of amphibians and of reptiles. We consider that such dermal structures have arisen many times in diverse lineages of vertebrates, and that these are expressions of properties of dermal collagen to support mineralization by special- ized dermal cells. However, we recommend that the term “dermal scale” be used for the mineralized dermal units of osteichthyans and gymnophiones, and “osteoderm” for the dermal structures of frogs and squamates, with the understanding that the terminology recognizes certain convergent attributes of shape and structure, but not of process. The presence of dermal scales that are com- convergent evolution, in response to the struc- posed of an unmineralized base plate of collage- tural constraints of the tissue involved. nous fibers and a superficial layer of mineralized There have been several descriptions of the squamulae is unique to members of the amphib- scales of gymnophione taxa in the last 150 years. ian Order Gymnophiona among all living tetra- Mayer (1829a,b), Rathke (1852), and Leydig pods. Because gymnophione scales are structur- (1853) commented briefly on scale morphology. ally similar to those of teleost fishes, they have Sarasin and Sarasin (1887-90) described the been assumed to be homologous (Zylberberg et scales of Ichthyophis glutinosus from Sri Lanka, al. ’80; Ruibal and Shoemaker, ’84). However, and several subsequent analyses have corrobo- recent examinationof new data on gymnophione rated their conclusions about the relationship of scales, together with an assessment of the mor- the scales to other dermal and to epidermal phology of fish scales, the dermal ossificationsof structures. Cockerell (’11, ’12),Phisalix (’10, ’12), frogs, and the osteoderms of lizards, causes us to Datz (’23), Ochotorena (’32), Marcus (’34),Gabe question the assumption of homology of dermal (’71a,b), Casey and Lawson (’79), Zylberberg et scales. This analysis indicates that all dermal al. (’80), Perret (’82), and Fox (’83)described the ossificationsshare a number of features that are scales of diverse species, usually Ichthyophis or structural properties of the particular organiza- Hypogeophis, at various levels of resolution. Tay- tion of the dermis, and that the similarities and lor (’72) assembled an “atlas of scales” composed differences in scale and osteoderm structure of photographs and descriptions of single scales found in various lineages constitute evidence for from virtually all of the species of gymno- o 1990 WILEY-LISS, INC. 26 L. ZYLBERBERG AND M.H. WAKE phiones. He indicated that features such as scale and Microcaecilia unicolor from Guyana were size, shape, numbers and distribution, and pat- processed for analysis by several techniques. The tern of mineralization might be useful for system- specimens of Dermophis mexicanus from which atic analysis. However, Wake and Nygren ('87) scales were taken are deposited in the collections demonstrated marked individual, ontogenetic, of the Museum of Vertebrate Zoology, Univer- and sexual variation in nearly all scale parame- sity of California, Berkeley, and the M. unicolor ters for Dermophis mexicanus from a single in the Museum National d'Histoire naturelle, population, and suggested that scales are of little Paris. Scales were stripped from the annuli of use as taxonomic characters to diagnose gymno- several formalin-ked, alcohol-preserved speci- phione species. Although Feuer ('62) proposed mens of different sizes (and presumably ages) that scales might provide information about ages and sexes; all scales per annulus were stained of individuals, Zylberberg et al. ('80) presented with alizarin red-S (see Wake and Nygren, '87, evidence to the contrary. for details). Quadrants of skin including several The greatest numbers of scales occur in taxa annuli were fixed in neutral buffered formalin or of gymnophiones considered primitive based on in Bouin's mixture, embedded in paraffin, and other characters (osteology, myology, reproduc- sectioned sagittally or frontally. Stains and his- tive biology; Ichthyophis, Caudacaecilia, Rhina- tochemical reactions are summarized in Tables 1 trerna, Epicrionops-see Taylor, '68, '72; Nuss- and 2. Scales were prepared for scanning elec- baum and Wilkinson, '89). Scales are embedded tron microscopy by excising, mounting on stubs, in virtually all of the annuli (i.e., body rings) of critical-point drying, and sputter-coating with primitive species, and these taxa have several gold or gold-palladium alloy. Scales were exam- annuli per body segment. In gymnophiones, there ined and photographed with an ISI-DS-130dual- is a trend toward the reduction of numbers of stage scanning electron microscope or with a annuli and, concomitantly,of scales. Taxa consid- JEOL 8s M 35 SEM at an operating voltage of ered highly derived based on other characters 25 kV. (e.g., typhlonectids and scolecomorphids)charac- For transmission electron microscopy, small teristically lack secondary annuli and scales (though Wake, ['75] and Moodie, ['78] found pieces of skin were removed from several regions scales in some Typhlonectes). of the body, fixed in a mixture of 2.5 % glutaral- However, the scales of few taxa have been dehyde and 2 % paraformaldehyde in 0.1 M ca- examined in detail with comparison to out- codylate buffer (pH 7.4) for 2 or 3 hr at room groups in mind. We present new data on the temperature; specimens were washed in 0.1 M morphology of scales of Dermophis mexicanus cacodylate buffer containing 10% sucrose, and and Microcaecilia unicolor (both Gymnophiona: postfixed in 1%osmium tetroxide in 0.1 M ca- Caeciliaidae wde Nussbaum and Wilkinson, '891) codylate buffer. Some samples were decalcified to augment the detailed data available for Ich- in the fixative with 0.1 M ethylene-diamino-tetra- thyophis kohtaoensis (Gabe, '71a,b; Zylberberg acetic acid (EDTA) added for 1or 2 days at 4"C, et al., '80; Fox, '83) and Hypogeophis rostratus and then washed and postfixed. Ruthenium red (Casey and Lawson, '79; Zylberberg et al., '80). (Martino et al., '79) was used as an electron The morphology of the scales of D. mexicanus microscopic stain for extracellular polyanions and M. unicolor is compared with that of those (Luft, '71a,b) such as acid mucosubstances and scales described previously, and with that of acid phospholipids. Ruthenium red was added dermal ossifications in other vertebrates. Our to the fixativecontaining EDTA, the wash buffer, comparative morphological approach elucidates and the osmium tetroxide solution. All the fixed the similarities and differences in mineralization samples were dehydrated in ethanol and cleared patterns of such dermal derivatives as scales and in 1-2 epoxy-propane. They were left in a 1:l osteoderms and allows assessment of homology epoxy-propanernpon mixture overnight or versus convergence of teleost scales, caecilian longer, and then transferred to fresh resin. Poly- scales, and anuran and reptilian osteoderms. The merization was carried out at 60°C. Thick sec- study also provides a basis for continued re- tions (-1 pm) were stained with toluidin blue for search into the questions of development and light microscopy. Thin sections of selected areas homologies of dermal structures in vertebrates. were cut with a Reichert ultramicrotome using a diamond knife with the tissue block oriented obliquely to the knife to improve sectioning of MATERIALS AND METHODS hard fibrous tissues (AUizard and Zylberberg, Scales of Dermophis mexicanus from San '82). Sections were mounted on collodion-coated Marcos Prov., Guatemala, and Chiapas, Mexico, grids and stained with uranyl acetate and lead GYMNOPHIONE SCALE MORPHOLOGY 27 TABLE 1. Histologicalstains employed for analysis of scales Histological stain Reference Azan Heidenhain in Gabe ('68) Mallory's azan Humason ('79) Hematoxylin & eosin Humason ('79) One-step trichrome Gabe and Martoja in Gabe ('68) Histochemical reactions Specificity Reference Proteins Danielli's coupled tetrazolium reaction General reaction Gabe ('68) Ferric ferricyanidereaction